Implementation of algorithms based on digital image processing and support vector machine to characterize the establishment of Tifton 85 and Jiggs grass pastures from RPAS images

• Grant number: 22/10653-0
• Support Opportunities: Scholarships in Brazil - Master
• Start date: June 01, 2023
• End date: February 29, 2024
• Field of knowledge: Agronomical Sciences - Animal Husbandry - Pastures and Forage Crops
• Principal Investigator: Carlos Guilherme Silveira Pedreira
• Grantee: Rigles Maia Coelho
• Host Institution: Escola Superior de Agricultura Luiz de Queiroz (ESALQ). Universidade de São Paulo (USP). Piracicaba , SP, Brazil

Abstract

The objective of the present work is to characterize the establishment of Tifton 85 and Jiggs grasses (Cynodon spp.) through the digital processing of images obtained with RPAS (Remotely Piloted Aircraft System) and, in parallel, apply an algorithm based on Support Vector Machine aiming at automatic classification. of land occupation in the initial phase of pasture formation. The project will be carried out in an experimental area located at Embrapa Pecuária Sudeste in São Carlos, SP, according to a randomized complete block design, with Tifton 85 and Jiggs grasses as treatments, in 8 replications. The establishment dynamics will be characterized by destructive sampling and by RPAS. The flights will be carried out weekly after the implantation of the cultivars in the field. Image collection will take place with the Mavic Enterprise Dual platform. The images collected weekly are orthorectified and submitted to routines based on digital image processing (PDI) elaborated in python, the values extracted from the images will be correlated to sampling and destructive methods in the experimental units through a regression analysis. The support vector machine (SVM) will be implemented with the objective of characterizing the vegetation cover of the plots, as well as differentiating the species of interest. The algorithm will be tested in different configurations in order to determine the most efficient model in supervised classifications.

Academic Publications

(References retrieved automatically from State of São Paulo Research Institutions)

COELHO, Rigles Maia. Characterization of establishment and forage mass prediction of Cynodon spp. pastures through remote sensing. 2024. Master's Dissertation - Universidade de São Paulo (USP). Escola Superior de Agricultura Luiz de Queiroz (ESALA/BC) Piracicaba.